The Financial Conduct Authority (FCA) is launching a “supercharged” digital sandbox that provides greater computing power, enriched datasets, and more sophisticated tooling, so firms can build their early-stage proof of concept models with support from the regulator.
“Working with our Sandbox partners, NayaOne, we will be bringing Nvidia tooling and resources to users of the Supercharged Sandbox – accelerating AI innovation and development,” said Jessica Rusu, chief data, information and intelligence officer at the FCA, in a speech. “The Supercharged Sandbox significantly upgrades our existing Sandbox programme, offering advanced Graphics Processing Unit (GPU) compute resource, enabling users to efficiently train, refine, and experiment with cutting-edge AI models.”
Participating firms will have access to Nvidia AI Enterprise software suite, a cloud-native suite of software tools, libraries, and frameworks, that accelerate and simplify the development, deployment, and scaling of AI applications.
In early June, the FCA also released a report examining the potential usefulness and limitations of large language models (LLMs) such as OpenAI’s GPT series in consumer-facing financial services.
To explore practical applications, the regulator conducted 2 pilot projects:
- Simplifying financial concepts: asked GPT-3.5 and GPT-4 to generate simplified definitions of complex financial terms, tailored to lower reading ages and supported by relevant examples.
- Providing consumer guidance: compared the effectiveness of LLM-generated responses in a fixed chatbot for cash savings queries with a traditional website-based Q&A format.
The 3 main lessons were:
- LLMs demonstrate strong potential in simplifying complex information, enhancing readability and accessibility. However, validating their outputs requires a robust evaluation framework that combines human judgement with automated tools.
- The effectiveness of LLMs is context-dependent. Outcomes such as user comprehension and engagement are influenced by how the model is embedded within the customer journey, including content design and delivery.
- There is a clear appetite for AI-driven assistance. Many users responded positively to automated support, indicating a readiness to engage with intelligent systems in decision-making processes.